The activity describes a community that starts with Seed as a general-purpose knowledge environment, gradually structures its own knowledge, and eventually extends the system to fit its own practices.

Knowledge community → Semantic network → Ontology → Tools → Applications → Community hacks its own system

1. Bootstrap the community’s knowledge space

As a community organizer, I want to create an initial shared space with a purpose, participants, and starting documents, so that the community has a place to begin working together.

As a community member, I want to bring existing documents, links, and references into the space, so that we can start from the knowledge we already have.

As a community member, I want to create and link concepts as they emerge, so that the community can progressively build a shared semantic network.

2. Develop a shared knowledge model

As a community member, I want to define concepts and their relationships, so that we can make our shared understanding more explicit.

As a knowledge organizer, I want to introduce attributes, types, and constraints when recurring structures emerge, so that informal knowledge can evolve toward a formal ontology.

As a community member, I want documents and objects to use the community’s ontology, so that knowledge can be consistently described, queried, and reused.

3. Turn recurring work into tools

As a community member, I want to identify repetitive workflows that could be automated, so that we can turn community practices into reusable tools.

As a developer or agent, I want to create a tool that reads and modifies Seed objects using the community’s knowledge model, so that software can operate directly on the community’s knowledge.

As a community member, I want to install and use tools created by people I trust, so that I can extend my system without giving arbitrary code unrestricted authority.

4. Hack and evolve the system itself

As an advanced community member, I want to modify or create applications on top of Seed objects, so that the interface can adapt to how our community actually works.

As a community member, I want to inspect who created a tool, what code it runs, and what permissions it requires, so that I can decide whether I trust it.

As a community, I want successful experiments to become shared applications and practices, so that the system evolves together with the knowledge and workflows of the community.

That makes the activity feel less like “customization” and more like progressive end-user programming: the community starts by expressing knowledge and eventually turns that knowledge into executable behavior.